Storage Migration Service makes it easy to migrate one or more servers to newer hardware or virtual machines. Hi, I’m Jeff Woolsey and this is one in a series of videos that shows you what’s new in Windows Admin Center. Windows Admin Center allows you to streamline everyday tasks and connect to Azure for hybrid scenarios. In this video, Ned Pyle shows you how the Storage Migration Service works. From Windows Admin Center, you can install the service, easily inventory multiple servers and their data, and validate that your destination servers meet the requirements. The Storage Migration Service uses an easy step-by-step workflow that walks you through the entire migration process, maps your legacy file server to the new one, and ensuring everything such as data, metadata, ACLs, and more, are all successfully migrated. In the final step, you’ll see how the File Server identity is moved from the source server to the destination server, so the change is invisible to users and avoids disrupting your apps. Let’s get started. Ned: We created a tool in 2019, called SMS to take care of it. It moves all your data. Anywhere you want. It’s pretty fast, it’s very consistent. High fidelity. It’s very scalable. It takes care of all the complexity and all the weird things you might not have thought about, like servers with alternate computer names and weird share permissions. It has a really nice Windows Admin Center workflow. And so you have all these source machines, even 2003. Yes, I have you covered for 2003. You add an orchestrator. It copies all this data to your regular servers, your Azure File Sync servers, physical servers, IS server…. I don’t care. You’ll get them there. In August, if you were paying attention, we released a very large cumulative update called the 8C release. And it… being put onto your orchestrator, added support for Samba and clusters… …for network-less migrations. And what does it do? You take an inventory. You transfer the data. You take over all the config. And you’re done. We take over the identity of the server. This cumulative update added Samba-Linux support for sources, cluster support for sources and destinations, including I have a standalone server that I want to convert into cluster. I mean, if that server was that important it should be clustered. Local group and user migration. Your robocopy scripts are not getting that, right? You’re losing all of that security every time you copy files between servers. We will put it all there for you, update the ACL, stitch it all together. And network-less migration. Inventory is passive. There’s no agent. You run it from the orchestrator. You just create a job. You create a name for it, you inventory the shares. We scan the volume, figure out, you know, how much things need to be copied, let you decide later which things you do and do not want to copy, and we’re ready to go for transfer. What does this look like? So I have a 7-minute demo… total… of SMS running from a couple 2008 and R2 servers, 2012 servers and 2008 servers, up until 2019 servers…. If you don’t have SMS installed, we will install it for you. We create a job. Type in whatever you like. We’re then going to point ourselves to one or more servers. You can point it up to 100 servers. I mean that’s kind of silly, but most people usually just point to one, but you can point to lots of servers. You can create lots of jobs. You can run them in parallel. You can have lots of orchestrators. No big deal. I’m going to log in as my Meghan the migrator account. Put in a password that is going to work as an administrator on my source machines. And I am just going to take a look at this machine that I plug in. Now, if you used SMS in the past, surprise #2. We finally added Active Directory browsing, so you can select servers, search for servers, not have to type in server names, it’s just like 1999. It’s really great. Windows Admin Center has come a long way. It’s almost as good as NT now. I could also just type in server names. Go ahead and add those. And if I choose “Start scan” right now, it will go through and inventory these machines, looking for every single file, count everything up, look at the storage, you know, give me an idea of what the security situation looks like. All these nice little details. Tell me the OS, these are all my shares. I seem to have some game shares. Configuration, network adapters, volumes, because we’re going to steal all this stuff. We’re going to subsume that machine into another one. And if I look at the actual source machine here, you can see, those are the same shares, right? My doom shareware share. And they have all kinds of weird permissions, and stuff on them and I’m going to get all of those types of weird permissions off of the shares and the NTFS permissions. And the encrypted files. And all the weird stuff people tuck away into various shadowy parts of IT. Like right here, I have on this particular file system, a local account. Eee. What am I going to do about those? That’s the local IT staff account. Somebody was like an NT 4 MCSE over there. OK. Now I’m also going to do inside my 7 minute demo, a cluster and a Samba demo. Alright so here. There’s my cluster demo. It’s the same. You just run the same thing, it’s fine. We just wrote cluster on there, it works. It’s no difference. What about the Samba-Linux demo? Alright there it is. Yeah, we have a screen we put in all the Linux creds. That’s it. That’s the rest of the whole demo, everything else works the same. It’s pretty nice right, like the whole process you don’t have to relearn. So I’m going to do transfer. I’m going to copy all the data, copy all the security. It’s multithreaded. It’s multi server. We will retry copies. We will copy only the differences. We will map everything. So what you’re going to do is pick a destination server or servers. Match up which things–you know–which server goes to which one. It will validate that your configs looks good and then we’re going to go ahead and run the transfer and be ready for cutover. So this is still part of my 7-minute demo. And so now I’m running into transfer. I’m going to give it some more creds because in case the creds in the source aren’t as good as the–you know– to run on the destination. So you don’t have to type in server names anymore. And wild card support. Yeah. And I scanned these things and it’s going to look at my destination and make sure that I have enough storage, that there is storage, that it’s formatted, that there’s drives. All that kind of good stuff. And I get my chance to decide if I want to include some things in my transfer or not. Typically, people don’t want to turn anything off. But maybe you have some share that’s just full of junk. And this is your chance to finally leave it in the dust. Notice how I always say transfer. I’m talking about copying data, not moving data. We will not touch your source machine. It’s basically a perfect back up at all times. So no matter how badly things go in your migration, no matter how badly your destination ends up looking, no matter how torqued off it is, your source is still there, even after cutover. We will still have your source machine online, just not available to users and apps anymore. Just available to you. So now I can just transfer settings. Here I get my chance to pick out local users and groups, or not to pick them out, or to change the behavior of whatever happens if I collide the names. And then I can do my favorite, really my absolute favorite part of SMS is, instead of reading a document about the requirements, I can press a button about the requirements. And if I don’t meet them, I don’t need to keep clicking Next in order to go find out later that it’s not going to work. Everything that needed to happen is right here. All right there. If I don’t pass those, I don’t need to collect $200 or go. OK, let’s start the transfer. And I’m not time-compressing this part. My 7-minute demo works because I’m not going to copy very much data. So, your transfer will not be 7 minutes. It’ll be like, however long it takes to copy 80 trillion zillion bytes of files. I’m kind of cheating. Surprise #3. We finally added in, not only progress throughput, but a time estimator of when it will be done. Yes. And like all progress bars, it’s totally full of it, so we’ll do our best estimate based on the current throughput, and then somebody will choke your network out and we will be like, “Oh no, come back tomorrow.” OK, so that’s how transfer looks. Now, this last thing is cutover so we’re going to take over the server–the old server’s name, its AD account, its IP addresses, its DNS, all that kind of crazy crap. And it’s going to be like nothing ever happened. We are going to take over and steal the identity and config of that source machine, plunk it onto the destination and then make the old source machine have some new scrambled name that either you set or we randomly generate, some new IP address that you set or you DHCP, whatever you want. So, your old server’s still there, you can still get to it no matter what. Your users can’t and your apps can’t unless they’re mind readers. OK? So we can set our cutover parameters, which are very, very, very few basically you’re just saying how long is too long for cutover to take. We still do validate and then we’re done. So, this is the last of my 7-minute demo. So now I’m into the cutover phase. All my credentials are saved from previously. If you want to change them in the meantime, you can. I can tell the destination server, OK, this NIC on the source, put your IPs on this NIC on the destination. If you got two or three, line them up. If you don’t want to copy all of them– and you should–but if you got IP addresses they don’t want to use anymore, you can do partial network migrations where you skip some NICs. We try to bring along IP addresses because everybody here has some pinhead developer, who wrote IP addresses into a script or an app or something somewhere, right? And today, this morning, we announced Azure Extended Network. Did you hear about that? Layer 2 switching into your Azure tenant. Use your on-prem IP addresses in Azure. Over Express or VPN. Go back and look for Greg Cusanza’s presentation for this morning and watch it. It’s amazing… no one has that. So now I’ve done my validation. It’s going to pass, because this is a demo. And I’m going to start the cutover. This is telling me all the things need to happen, it’s really telling you how cutover works. So I’m going to choose Next. I’m going to start the cutover. I click it. And because I want to have another 7 minutes, you know, my demo, to last 7 minutes, I’m going to…cheat. two servers to reboot and your DNS to finally catch up with the universe with replication is how long your cutover will take. And that’s how I make a 7-minute demo of SMS.